A main bottleneck for realizing green-emitting semiconductor lasers and other high Indium containing InGaN optoelectronic devices is the lack of a suitable lattice-matched template for growing the InGaN heterostructures needed to produce the target device. Typical green-emitting semiconductor lasers utilize quantum wells made from InGaN containing more than 25% Indium. Conventional methods for producing such high Indium containing InGaN optoelectronic devices include utilizing GaN or GaN on sapphire templates. However, the use of these conventional templates is inherently problematic because of the large lattice mismatch between GaN and the target InGaN heterostructures, which leads to high strain and unacceptable levels of performance-degrading material defects and built-in polarization fields.
More recently, efforts have been made to reduce GaN-InGaN mismatch defects by utilizing strain reduction superlattice layers between the GaN template and the target InGaN composition. Unfortunately, these efforts have not been successful in reducing defects to a satisfactory level.
Attempts to produce high-Indium InGaN heterostructures using conventional techniques have proven inadequate because the resulting InGaN layer is either too thin or too rough. In particular, it has been observed that InGaN grown on GaN using conventional methods undergoes some kind of phase transition above about 100 nm (nanometers). For example, a 100 nm In0.10Ga0.90N film grown on GaN using conventional methods appears excellent, but the film becomes very rough and develops multiple x-ray peaks when grown just twice as thick. It is believed that this problem arises due to a natural immiscibility of the InGaN alloy, which limits the film thickness to about 100 nm for films containing about 10% indium. The maximum film thickness may be even thinner for the higher indium content required in green laser diodes.
The immiscibility problem is likely not fundamental to InGaN. Rather, it is caused by strain when trying to grow high indium-containing InGaN on GaN. Therefore, the immiscibility issue can be resolved, and a thick high quality high-indium containing InGaN can be achieved by growing the film on an InGaN on sapphire template (where strain would be small), instead of on a GaN template (where strain would be high).
This invention is directed toward structures and methods for attaining devices on InGaN templates. It is important for the template to be relaxed and unstrained because its key function is to provide a new lattice parameter for growing high Indium-containing heterostructures.
A green-emitting semiconductor laser made from InGaN contains more than 25% Indium in its active region. A suitable InGaN template for that device requires at least about 10% Indium in the uppermost layer in order to produce acceptable levels of strain similar to those present in available blue and blue-violet lasers. It is also anticipated that a suitable InGaN template needs to be reasonably thick—perhaps about 3 to 5 μm (microns) thick—in order to produce excellent structural and surface qualities.
What is needed is method for reliably forming relaxed InGaN templates for high-Indium InGaN heteroepitaxy that address the problems described above, and to the relaxed InGaN templates formed by such a method.